Published online before print September 18, 2003, 10.1148/radiol.2292020982
(Radiology 2003;229:598-602.)
© RSNA, 2003
Interventional MR Angiography with a Floating Table1
Harald H. Quick, PhD,
Hilmar Kuehl, MD,
Gernot Kaiser, MD,
Stephanie Aker, DVM,
Silke Bosk, RT,
Jörg F. Debatin, MD and
Mark E. Ladd, PhD
1 From the Departments of Diagnostic and Interventional Radiology (H.H.Q., H.K., S.B., J.F.D., M.E.L.), Transplantation Surgery (G.K.), and Pathophysiology (S.A.), University Hospital Essen, Hufelandstr 55, D-45122 Essen, Germany. Received August 5, 2002; revision requested October 8; revision received January 29, 2003; accepted March 11. Supported by DFG grant no. LA 1325/1–1. Address correspondence to H.H.Q. (e-mail: hhquick@uni-essen.de).
 |
ABSTRACT
|
|---|
A floating table was integrated into a setup for performance of interventional magnetic resonance (MR) angiography procedures with actively visualized catheters and biplanar real-time image fusion. The setup was evaluated by performing catheterizations in eight pigs. The floating table enabled the authors to follow actively visualized instruments in the pigs’ vasculature during MR imaging–guided interventional angiography procedures while performing real-time biplanar MR imaging. Interventional MR angiography with a floating table enables the field of view to be moved along with the instrument tip to the region of interest and thus enhances the usability and flexibility of the interventional MR imaging setup.
© RSNA, 2003
Index terms: Angiography, technology, **.1242 • Animals • Aorta, interventional procedures, **.124, **.126 • Interventional procedures, experimental studies, **.124 • Magnetic resonance (MR), vascular studies, **.12142
|
INTRODUCTION
|
|---|
Recent advances in the field of interventional magnetic resonance (MR) angiography have propelled it to become an increasingly promising research frontier. Active instrument visualization techniques (1–6) allow for sequence-independent visualization of multiple vascular instruments with high contrast. Fast steady-state imaging techniques, such as true fast imaging with steady-state precession (FISP) (7), allow for real-time imaging with high spatial and temporal resolution and excellent contrast characteristics, rendering vessels hyperintense in all imaging orientations, even without administration of a contrast agent (8,9). Fast reconstruction and image fusion techniques enable real-time image reconstruction, postprocessing, and fusion, and thus simultaneous visualization of instruments and anatomy (9).
To date, MR imaging in general and interventional MR angiography procedures in particular have been limited to a single region of interest. Spatial constraints of the imaging volume, which are determined on the basis of magnetic field homogeneity and gradient linearity, limit the maximum achievable field of view (FOV) and thus the region to be imaged. Vascular instrument guidance, however, requires the instrument to be imaged from its introduction into the vasculature to the therapeutic region of interest, which often requires the instrument to be advanced beyond the spatial constraints of the imaging volume. To follow the instrument from the port of first entry to its final destination within the vasculature, it is thus mandatory to reposition the patient within the isocenter of the magnet, along with any signal-receiving radiofrequency (RF) surface coils that are placed on top of the patient to enhance image quality. This procedure is cumbersome, unpractical, and time consuming, however, and thus limits the overall flexibility of performing MR imaging–guided interventional vascular procedures.
To extend the longitudinal FOV, a variety of methods have been developed (10–14) in which two- or three-dimensional acquisitions are performed at each of several discrete (fixed) stations, with the FOV of each station being equal to or smaller than the largest imaging volume possible, given the magnetic field homogeneity and gradient linearity. Some of these methods involve the use of a floating table that can be repositioned manually relative to the isocenter of the magnet. For example, floating table techniques allow for whole-body MR imaging applications, such as whole-body MR angiography (11,12) and whole-body screening for bone metastases (10,14). Other methods have been developed that extend the longitudinal FOV without discrete stations by allowing continuous data acquisition during table movement (15,16).
The purpose of the present study was to incorporate a floating table into an existing experimental setup (9) to perform vascular interventions with actively visualized catheters and biplanar real-time MR image fusion.
|
Materials and Methods
|
|---|
Interventional Real-Time Imaging Setup
MR imaging was performed with a 1.5-T Sonata system (Siemens Medical Solutions, Erlangen, Germany) equipped with eight RF receivers, each with plus or minus 250-kHz maximum bandwidth and a gradient system capable of 40 mT/m maximum amplitude and 200 mT/m/msec slew rate. Two elements from the anteriorly placed body phased-array coil (Body Array Flex; Siemens) and the posteriorly located spine phased-array coil (Spinal array; Siemens) were used to image tissue morphology.
To enable active and independent visualization of the vascular instruments, a 0.035-inch guide wire and 6-F catheter were designed to contain electric dipole antennas (Somatex Medical Devices, Berlin, Germany) (3,5,9). The instruments were connected to individual RF-shielded boxes that contained the requisite electronics for antenna tuning, matching, and active decoupling during RF transmission with the whole-body RF coil of the MR imager. The boxes were connected to individual receive-only surface coil ports of the MR imager. The guide wire and catheter were 130 and 105 cm long, respectively.
The MR imaging sequence for instrument guidance was a real-time true FISP sequence with a repetition time msec/echo time msec of 2.2/1.1, flip angle of 50°, FOV of 30 x 30 cm, matrix of 256 x 135, section thickness of 10 mm, and bandwidth of 1,085 Hz/pixel. The acquisition time per image in monoplanar imaging mode was 297 msec, resulting in a frame rate of 3.4 images per second without any echo sharing. For biplanar instrument guidance, the true FISP sequence was modified to allow for optional interleaved acquisition of two orthogonal imaging planes. The image acquisition frame rate for each plane was thus reduced by a factor of two.
By implementing special software on the standard reconstruction computer of the MR imager, the individual receiver channels could be identified and combined separately so that individual images of the guide wire, catheter, and anatomy were reconstructed and then transferred to an image fusion personal computer (1.7-GHz Pentium IV processor with 1 GB random access memory) via a 100 Mb/sec Ethernet connection. Software implemented in Java (PMOD, www.pmod.com) enabled individual windowing, leveling, color coding, and subsequent fusing of the real-time images. The fused composite images were displayed on a 13-inch TFT monitor (in-room console, Siemens Medical Solutions) in the imaging room for viewing by the interventionalist. For biplanar visualization of the resultant images, the output display of the PMOD fusion program was split with a dual-monitor graphic card, which resulted in interleaved visualization of both imaging planes on two in-room TFT monitors (additional 17-inch TFT personal computer monitor). The complete imaging setup is described in more detail in the article of Quick et al (9).
Floating Table Platform
A manually positioned floating table platform (AngioSURF; MR Innovation, Essen, Germany) was used to move the pig in the longitudinal (head-to-feet) direction on top of the original table of the MR imager. This system was originally developed for performance of contrast material–enhanced multistation whole-body MR angiography (11). The table covers a longitudinal FOV of up to 200 cm and allows for continuous or discrete multistation movement. For the pig experiments, the table could be used without further modification simply by leaving out the rastering mechanism that is used to define the stops for the discrete stations. The maximum table speed during the in vivo experiments was approximately 3 cm/sec.
The floating table platform system consists of two roller boards, a coil glider to support the body-array surface flex coil (Body Array Flex; Siemens), and a table platform that can be moved manually in the longitudinal direction on top of the imaging table (Fig 1). The coil glider with the anterior body-array flex coil is held in its position at the isocenter of the magnet by two support rods that are connected to the foot end roller board. Thus, the anterior two-element body flex surface coil is always positioned in the isocenter and follows the contours of the pig’s anatomy during table movement. Two elements of the spinal-array coil integrated into the imager table (Spinal array; Siemens) provided signal detection from the posterior regions of the animal. This setup ensures optimal proximity of the surface coils to the regions of the animal to be imaged and thus combines the high signal-to-noise ratio of surface-array coils with an extended longitudinal FOV.
View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Floating table system mounted on top of the MR imaging table, which contains the built-in spinal-array (SA) RF surface coil. Imaging table cushions superior and inferior to the spinal-array surface coil are replaced by two roller boards (RB). The small roller board is connected to the coil glider (CG), which rests against the opening of the magnet bore. The coil glider hosts the body-array surface RF coil and centers the coil within the isocenter of the magnet. For imaging, a floating table is placed on top of the roller boards.
|
|
In Vivo Experiments
In vivo experiments were performed with fully anesthetized domestic pigs (n = 8) weighing 40–60 kg. The experiments were conducted in accordance with all regulations set forth by institutional and governmental agencies. The animals received 5,000 IU of heparin 2 hours prior to the intervention. A 7-F introducer sheath was inserted into the right iliac artery by means of arterial cut-down in open surgery. The animals were placed head-first and supine on top of the floating table with the legs and ventilation tube secured with gauze to the table platform (Fig 2). The guide wire and catheter were inserted through the introducer sheath and guided through the vasculature by using real-time true FISP MR imaging in the biplanar imaging mode. While the interventionalist (H.K.) performed the instrument manipulations, another in-room operator (H.H.Q.) manually moved the floating table platform to follow the course of the instruments through the animals’ vasculature. Selective catheterization of the abdominal aorta, renal arteries, hepatic artery, and superior mesenteric artery was performed. After the experiments, the animals were euthanized with a bolus injection of 80 mg per kilogram of body weight of pentobarbital (Eutha 77; Essex Pharma, Munich, Germany).
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Pig on top of the floating table, head first and supine with the legs fixed with gauze and tape. The floating table allows the fully anesthetized and ventilated pig to be moved manually between the spinal-array surface coil and the phased-array body flex coil that is mounted on the coil glider. A-C, Image series demonstrates how the pig is moved to position both surface coils at a new region of interest. For demonstration purposes, photographs were taken outside the bore of the imager.
|
|
The practicability of the floating table within the interventional imaging setting was evaluated by comparing it with a conventional setting without a floating table, with the animal’s anatomy covered by multiple surface coils. Image quality was investigated with regard to signal homogeneity, image sharpness, and artifacts by means of independent review by two investigators (H.H.Q., M.E.L).
|
Results
|
|---|
Floating Table Platform
The floating table platform enabled easy movement of the animal in the longitudinal direction, which allowed us to follow the instruments through the pigs’ vasculature during the course of an intervention (Fig 3). The body-array flex coil mounted on top of the coil glider provided optimal adaptation of the proximity of the coil to the contour of the pigs’ anatomy, which provided continuous optimization of signal-to-noise ratio during table movement. Four active surface coil elements were sufficient to receive signal over the 300 x 300-mm FOV anteriorly to posteriorly. No relevant signal drop-off could be observed toward the edges of the FOV. Since the FOV was always centered around the homogeneous isocenter of the magnet, no imaging artifacts and distortions due to gradient nonlinearities could be detected, which resulted in excellent image quality. No imaging artifacts or blurring could be observed when the table moved during real-time imaging at a maximum speed of approximately 3 cm/sec. At higher table speeds, image blurring was observed, which resulted in decreased signal intensity and thus decreased image contrast, especially for signal coming from the vessel lumen.
View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Advancement of an actively visualized guide wire and catheter from, A, the iliac artery through, B, the abdominal aorta into, C, the aortic arch of a pig. The MR images were acquired in the coronal plane. The actively visualized guide wire is displayed in red, while the catheter is displayed in green and yellow. The colored instrument outlines are overlaid onto the anatomic images that were acquired with the body and spinal-array coils (A-C). D-F, The corresponding images in the bottom row show schematically how the catheter is advanced into the pig while the pig, lying on the floating table, is moved out of the imager. Arrowheads indicate the catheter tip. This procedure ensures that the moving region of interest always stays within the FOV that is covered by the body and spinal-array coils of the imager (black rectangles anterior and posterior to the pig). Imaging parameters were true FISP with 2.2/1.1, flip angle of 50°, FOV of 30 x 30 cm, matrix of 256 x 135, section width of 10 mm, bandwidth of 1,085 Hz/pixel, and frame rate of 3.4 images per second without any echo sharing.
|
|
In Vivo Experiments
The guide wires and catheters were successfully guided into the aorta and its major arterial branches—the renal arteries, aortic arch, hepatic artery, and superior mesenteric artery—in all pigs. When compared with the conventional setting without a floating table, the proposed setup allowed for continuous image acquisition during table movement, obviating the need for repositioning of the animal, coil, and/or FOV and associated time delays.
Biplanar MR imaging provided images of the abdominal aorta of pigs alternately in the coronal and sagittal planes, with a reduced frame rate of 1.7 frames per second per plane. Thus, movement of the floating table during biplanar image acquisition enabled us to follow the advancement of the instruments along the pig’s aorta in two independent planes (Fig 4).
View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Biplanar real-time MR images show the advancement of an actively visualized guide wire (red) and catheter (green and yellow) from the iliac artery through the abdominal aorta into the aortic arch of a pig. A-E, Images acquired in the coronal plane. F-J, Corresponding images acquired in the sagittal plane. Ten of 360 acquired frames are shown. As soon as the advancing instruments reached the top edge of the FOV (B and G), the pig was manually pulled out of the imager while the instruments were advanced further (C-E and H-J). This always kept the catheter tip within the visible FOV. Arrowheads mark the anatomic position of the diaphragm. Imaging parameters were identical to those in Figure 3, and frame rate was 1.7 images per second due to interleaved biplanar image acquisition.
|
|
|
Discussion
|
|---|
The floating table platform obviates the need to repeatedly reposition the animal to the isocenter of the magnet whenever movement of the FOV relative to the anatomy is necessary. Undesirable imaging interruptions are thus avoided. There is also no longer a need to cover the animal’s anatomy with multiple surface coils (which are each connected to multiple receivers) and to switch between individual coil elements. Multiple coils could potentially be used to provide whole-body surface coil coverage; however, limitations in the number of coil plugs may make this unachievable with current imager designs.
The floating table approach streamlines interventional procedures and provides whole-body surface coil quality with just two surface coils. Access to the animal is hindered only in the vicinity of the current FOV, and the surface coil can be quickly slid to a new position to provide universal access. Electrocardiographic and other monitoring cables can be placed under a low-friction cloth and led along the side of the animal without hindering the movement of the coil glider.
The proposed concept shows some limitations. For the real-time true FISP MR imaging sequence, the maximum table speed that could be used before image blurring became apparent was approximately 3 cm/sec. At higher table speeds, image blurring and loss of signal and thus contrast were observed. Two effects might account for this observation. First, at higher table speeds, the imaging frame rate of 3.4 images per second is not sufficient to keep up with (
freeze
) the moving structures. Second, the true FISP MR sequence requires a number of excitations to obtain a steady state and thus increased signal intensity coming from the vessel lumen. Movement of the structures to be imaged beyond a certain limit between excitations leads to an insufficient buildup of steady state and thus limited image contrast. While a higher imaging rate would be desirable, especially to resolve motion during cardiac interventions, the table speed that could be achieved without image blurring in this study appears to be sufficient to match the speed of the advancing vascular instruments for most conceivable vascular interventions.
The imaging setup currently works with a predefined section orientation (two predefined section orientations in biplanar mode). For sharp display of anatomic structures, the imaging section should be as thin as possible and is currently in the range of 10 mm. During the course of an intervention, the position of the pig relative to the imaging section is constantly changed with the aid of the floating table. This might result in partial or complete loss of catheter visibility whenever the instrument (or parts of it) leaves the imaging section. An interactive graphic user interface or automatic section repositioning (1,17) steered by the instrument position could potentially compensate for this effect. Results of the current study, however, have shown that a coronal predefined section position is sufficient to visualize and thus guide the catheter along the aorta from its entry into the femoral artery up to the aortic arch, at least for aortas without substantial kinking.
The ideal suite for performance of MR imaging–guided vascular interventions should include, among other factors, a dedicated magnet that enables improved access to the patient during the course of the intervention. While low-field-strength open magnet designs allow for enhanced patient access and simple and robust MR imaging guidance of certain interventions, such as biopsy, these imagers show some limitations with regard to image quality and image acquisition speed. Because of the inherent limitation in B0 field strength, as well as limitations in gradient amplitude and slew rate, current open magnet designs do not appear suitable for guidance of vascular interventions with MR imaging. Here, conventional closed-bore designs have intrinsic advantages with regard to signal-to-noise ratio and imaging speed, which has motivated several investigators to explore the potential of performing interventional MR angiography with closed-bore 1.5-T systems.
The conventional closed-bore design, however, restricts convenient patient access during the course of an intervention. The current generation of 1.5-T closed-bore imagers has bore lengths of about 1.6 m, but there seems to be a trend toward even shorter magnets that will potentially provide enhanced patient access. The tradeoff for such short magnets, however, might be a further reduction in the homogeneous volume of the static magnetic field B0 and thus a limitation in the usable FOV. The floating table approach can potentially compensate for this deficit, since a homogeneous imaging volume with a longitudinal extent of about 30 cm already appeared sufficient for instrument guidance in these initial experiments.In conclusion, implementation of a floating table for performance of MR imaging–guided vascular interventions with actively visualized catheters and real-time biplanar image fusion enhanced the usability and flexibility of the experimental setup.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank Hongwei Zhang, MD, for technical assistance in performing the animal experiments.
|
FOOTNOTES
|
|---|
2 **. Multiple body systems 
H.H.Q., J.F.D., and M.E.L. are founders of MR-Innovation, Essen, Germany.
Abbreviations: FISP = fast imaging with steady-state precession,
FOV = field of view,
RF = radiofrequency
Author contributions: Guarantors of integrity of entire study, H.H.Q., M.E.L.; study concepts and design, all authors; literature research, H.H.Q., M.E.L.; experimental studies, H.H.Q., H.K., G.K., S.A., S.B., M.E.L.; data acquisition, H.H.Q., H.K., S.B., M.E.L.; data analysis/interpretation, H.H.Q., H.K., M.E.L.; statistical analysis, H.H.Q., M.E.L.; manuscript preparation, H.H.Q.; manuscript definition of intellectual content, editing, revision/review, and final version approval, H.H.Q., J.F.D., M.E.L.
|
REFERENCES
|
|---|
- Dumoulin CL, Souza SP, Darrow RD. Real-time position monitoring of invasive devices using magnetic resonance. Magn Reson Med 1993; 29:411-415.[Medline]
- Leung DA, Debatin JF, Wildermuth S, et al. Intravascular MR tracking catheter: preliminary experimental evaluation. AJR Am J Roentgenol 1995; 164:1265-1270.[Abstract/Free Full Text]
- McKinnon GC, Debatin JF, Leung DA, Wildermuth S, Holtz DJ, von Schulthess GK. Towards active guidewire visualization in interventional magnetic resonance imaging. MAGMA 1996; 4:13-18.
- Glowinski A, Adam G, Bucker A, Neuerburg J, van Vaals JJ, Gunther RW. Catheter visualization using locally induced, actively controlled field inhomogeneities. Magn Reson Med 1997; 38:253-258.[Medline]
- Ocali O, Atalar E. Intravascular magnetic resonance imaging using a loopless catheter antenna. Magn Reson Med 1997; 37:112-118.[Medline]
- Ladd ME, Zimmermann GG, Quick HH, et al. Active MR visualization of a vascular guidewire in vivo. J Magn Reson Imaging 1998; 8:220-225.[Medline]
- Oppelt A. FISP: a new fast MRI sequence. Electromedica 1986; 54:15-18.
- Peters DC, Dick AJ, Raman VK. Active guidewire tracking with real-time undersampled projection reconstruction (abstr) In: Proceedings of the Tenth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 2271.
- Quick HH, Kuehl H, Kaiser G, et al. Interventional MRA using actively visualized catheters, TrueFISP, and real-time image fusion. Magn Reson Med 2003; 49:129-137.[CrossRef][Medline]
- Barkhausen J, Quick HH, Lauenstein T, et al. Whole-body MR imaging in 30 seconds with real-time true FISP and a continuously rolling table platform: feasibility study. Radiology 2001; 220:252-256.[Abstract/Free Full Text]
- Goyen M, Quick HH, Debatin JF, et al. Whole-body three-dimensional MR angiography with a rolling table platform: initial clinical experience. Radiology 2002; 224:270-277.[Abstract/Free Full Text]
- Shetty AN, Bis KG, Duerinckx AJ, Narra VR. Lower extremity MR angiography: universal retrofitting of high-field-strength systems with stepping kinematic imaging platforms—initial experience. Radiology 2002; 222:284-291.[Abstract/Free Full Text]
- Bock M, Volz S, Zuehlsdorff S, Umathum R, Semmler W. Whole-body MRI: a simple approach using automatic table movement and dedicated post-processing (abstr) In: Proceedings of the Tenth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 562.
- Lauenstein TC, Freudenberg LS, Goehde SC, et al. Whole-body MRI using a rolling table platform for the detection of bone metastases. Eur Radiol 2002; 12:2091- 2099.[Medline]
- Kruger DG, Riederer SJ, Grimm RC, Rossman PJ. Continuously moving table data acquisition method for long FOV contrast-enhanced MRA and whole-body MRI. Magn Reson Med 2002; 47:224- 231.[CrossRef][Medline]
- Fain SB, Du J, Browning FJ, Polzin JA, Block WF, Mistretta CA. Floating table isotropic projection imaging (FLIPR): a technique for fast, extended FOV, contrast enhanced MRA (abstr) In: Proceedings of the Tenth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 212.
- Zuehlsdorff S, Volz S, Thesen S, et al. Interactive catheter tracking automatic slice repositioning controlled by the catheter (abstr) In: Proceedings of the 10th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 2272.
TOP
ABSTRACT
INTRODUCTION
